Bug pylint 4326 #1183
Bug pylint 4326 #1183
Conversation
astroid/brain/brain_numpy_utils.py
Outdated
|
|
||
| return tuple(numpy.version.version.split(".")) | ||
| except ImportError: | ||
| return None |
There was a problem hiding this comment.
Shouldn't we return (0, 0, 0) ? We'd get TypeError: '>' not supported between instances of 'NoneType' and 'tuple' line 25 if we can't import numpy.
| @@ -341,6 +345,59 @@ def test_datetime_astype_return(self): | |||
| ), | |||
| ) | |||
|
|
|||
| @unittest.skipUnless( | |||
There was a problem hiding this comment.
I think we can add a small test to check that we're not crashing if numpy is'nt installed.
There was a problem hiding this comment.
@Pierre-Sassoulas i'm sorry but i don't see the point here. Can you explain a bit what you want to prevent?
There was a problem hiding this comment.
I think it would have detected the problem of the other remark (TypeError: '>' not supported between instances of 'NoneType' and 'tuple'), by having automated tests for it.
There was a problem hiding this comment.
@Pierre-Sassoulas i was thinking to add something alone the following lines:
@unittest.skipIf(HAS_NUMPY)
class NumpyBrainUtilsTest(unittest.TestCase):
def test_get_numpy_version_do_not_crash(self):
"""
Test that the function _get_numpy_version doesn't crash even if numpy is not installed
"""
self.assertEqual(_get_numpy_version(), ('0', '0', '0'))
Is it what you were thinking of? Or do you had in mind something more generic?
There was a problem hiding this comment.
Looks good, but we could even check what we get when parsing the code without numpy (that would execute the _get_numpy_version too, but also much more code).
cdce8p
added
the
pylint-tested
hippo91
force-pushed
the
bug_pylint_4326
branch
from
7ecd80f
to
61fb19f
Compare
Steps
Description
This PR adds type hints support inside
numpy's brains. This way subscripting of classes that inherit fromgenericorndarrayis possible.Type of Changes
Related Issue
Closes pylint-dev/pylint#4326